crseek

.. image:: https://travis-ci.org/DamLabResources/crseek.png?branch=master :target: https://travis-ci.org/DamLabResources/crseek

.. image:: https://coveralls.io/repos/github/DamLabResources/crseek/badge.svg?branch=master) :target: https://coveralls.io/github/DamLabResources/crseek?branch=master

crseek is a tool for designing complex CRISPR-Cas9 workflows using the popular SKlearn API.

Quick Start

For the impatient:

from Bio.Seq import Seq
from Bio.Alphabet import generic_dna, generic_rna

from crseek.preprocessing import OneHotTransformer
from crseek.estimators import CFDEstimator
from crseek.utils import cas_offinder

gRNA1 = Seq('ATGTTGAGTCAGTGAAGGTG', alphabet = generic_rna)
gRNA2 = Seq('ATGTTCAGTCAGTAAAGGTG', alphabet = generic_rna)

possible_hits = cas_offinder([gRNA1, gRNA2], 5, direc='/path/to/genome/files/')

cfd_est =  CFDEstimator.build_pipeline()

target_scores = cfd_est.predict_proba(possible_hits.values)

See the notebooks in the paper/ folder for more examples.

Features

Preprocessing

• crseek.preprocessing.MatchingTransformer
• crseek.preprocessing.OneHotTransformer

These modules are modeled after the preprocessing modules in Sklearn and are capable of comparing (spacer, target) pairs. The outputs of these modules can either be fed further into the crseek estimation tools or used for building one's own machine learning methods.

Estimation

• crseek.estimators.MismatchEstimator
• crseek.estimators.MITEstimator
• crseek.estimators.CFDEstimator
• crseek.estimators.KineticEstimator

These modules can take the outputs of preprocessed (spacer, target) pairs and estimate the likelihood of cleavage under numerous assumptions. These modules implement .fit(), .predict(), and .predict_proba() methods and are usable in all Sklearn tools.

Utilities

• crseek.utils.smrt_seq_convert
• crseek.utils.extract_possible_targets
• crseek.utils.tile_seqrecord
• crseek.utils.cas_offinder

There is also a collection of utilities for manipulating sequence types, finding potential targets in sequences, and searching long (potentially genome scale) sequences for cutting events. This includes a wrapper around the popular cas_offinder tool for mismatch searches of large genomes using OpenCL-enabled devices. The searching methods have also been adapted to work with any potential PAM allowing the tool to be used with non-standard Cas9 variants.

Additional Features

• Tight integration with BioPython allows for the easy import, search, manipulation, and export of sequences.
• Integration with PySam allows for the search of SAM/BAM files.
• Preprocessing and Estimation tools intelligently deal with ambigious base calls in sequences allowing the tool to be used on draft-quality genomes.

Installation

Currently the installable modules are split between pip and conda tools due to multiple non-python dependencies.

conda config --add channels conda-forge
conda config --add channels defaults
conda config --add channels r
conda config --add channels bioconda
conda create -yq -n crseek-environment --file requirements.conda
source activate crseek-environment
pip install -q -r requirements.pip
python setup.py install